[OSM-dev] "The Future of Areas" - open discussion?

[Lennard]

On 25-4-2011 13:31, Jukka Rahkonen wrote:

>> also: which of the old data is most probably an area and which is a
>> closed linear feature. This might not be clear in all cases.

Not in all cases, but the overwhelming majority of them are clear.

> And further: There are some badly invalid area features in the database
> now because the multipolygon relation accepts anything. What would be done
> for those in the conversion?

Continue to serve such a case in their broken relation+members form. Let 
the editors write it back in that form if the user hasn't fixed it, the 
editors itself were unable to fix it, or upon notifying the user he/she 
could fix it but then chooses not to do so.

'Conversion by attrition' -- See: created_by=*. Of course, the large 
majority that are clear could be done in one fell swoop, in some kind of 
A-Day (Area Day, which in actuality would probably turn out to be an 
A-Weekend+ ).

-- 
Lennard